DR. ZIMRIN: Okay. Got it. Sorry; I should have read through this first. We have heard a lot of very interesting material today. I think that we know a lot more than we knew a couple of years ago. Its sounds like there are still a lot of questions that are unanswered at this point.

I am sure that there is more research that is planning to be done in this area, but I guess I am a little hesitant--maybe this is answering a question that is coming down later--to recommend universal testing at this point where there is so much that is up in the air. I would not, however--much as I would like the idea of questioning to bring out the people most at risk, it doesn't sound to me like that is a feasible approach given the inaccuracy with which people answer questions.

DR. DUNCAN: Well, you are right in one sense that our perception is that the questions would not be an adequate way of identifying the people at risk. But, keep in mind, that all the selective-testing strategies that we are seriously considering will follow one negative test. Everyone who comes in for the first to donate will be tested. And the issue of selective testing will be what happens when they come back again. Will they be asked questions to determine whether they are at risk? Will they be tested again, or will they not be tested again? Or will they be tested again every time they come?

DR. CRYER: So the answer to No. 1 is really moot. What we are really talking about is No. 2. Is that right?

DR. DUNCAN: What do you mean? There is no 2 up there yet. We want to get this one answered and then move on to 2.

DR. CRYER: But the question doesn't make sense because you are already--No. 2 already presupposes that everybody is going to get one test, as does this. What you are asking in No. 2 is--

DR. DUNCAN: No. 2 would be testing once with no questions asked. That is the difference between this and No. 2.

DR. NAKHASI: You have to take this question in its own right to say what you heard from the data presented whether the questioning is one of the viable options or not.

DR. Di BISCEGLIE: I would say, based on what we heard today, we heard no evidence that newly acquired infection is, indeed, a problem. It takes years to become infected living in an infested house. Reasonably casual travel doesn't seem to represent a risk factor. So it is not quite directly aimed at the question of whether questions are sufficient, but the general issue of newly acquired infection is a problem.

I did have a question for Dr. Forshee, though, which is somewhat related that I didn't get a chance to ask. In the modeling that you did, looking at test-once, test-twice, and universal, and you didn't really factor questions in. You said that. But you said that the difference between universal and test-twice, there is a numerical difference, that it didn't matter whether you factored newly acquired cases in or not. That didn't quite make sense to me. How could the number come out the same whether you did or didn't factor in newly acquired infection? Maybe I misunderstood.

DR. FORSHEE: Because the differences between those scenarios were based on the empirical sensitivity estimates that had been provided in Dr. Stramer's estimates. So our question there wasn't, how did the case get missed but, rather, just what percentage of cases would be missed under the alternative testing strategies of test-once, test-twice or use universal testing.

So, simply based on the empirical estimates of what the sensitivity was under those three estimates, we used those sensitivities to estimate the models and those were the differences that we found.

DR. ZIMRIN: So you are saying the autochthonous numbers are so low that it doesn't change the risk as compared to the other causes of positivity?

DR. FORSHEE: I'm sorry; could you repeat.

DR. ZIMRIN: Are you saying that the autochthonous are so low it is not going to change the risk estimate comparatively?

DR. FORSHEE: I am saying that the estimates in our model, to the extent that the autochthonous are in our model, they are in some of the cases that were missed in the test-once or test-twice strategy in the, I believe, it was 394 individuals who were looked at under the test-once and test-twice strategy. So, at this point, we didn't have sufficient data to sort out what those newly generated cases would be. So, instead, we relied on the estimates of sensitivity that were coming from the American Red Cross.

DR. EPSTEIN: If I could help shed some light on this. There are no accurate data, and really no data, to tell us how likely it is that a donor will acquire Chagas in the U.S. What Susan Montgomery showed you is it does happen. What we also saw from the cumulative testing by the blood organizations is that you will pick up autochthonous cases. However, what we are getting at is whether either additional testing or additional questioning could help us decide how much more testing to do in order to pick up newly emerged infections.

So what the survey found is that, over a 22-month observation period with an average 8 months per donor, per repeat donor, we didn't see an incident infections. Does that mean they don't occur? No; of course not. But do we believe that retesting donors based on a history of, say, camping or traveling south of the border, or, did you ever see a reduviid bug on your wall. We have no evidence that that would be of benefit.

Now, what I think is confusing about the question is we made the presumption that no one is going to advocate that you only test people based on a questionnaire history because the sensitivity of that methodology was low. It was 75 percent at best. So we presumed that there will be testing if there is intervention at all.

Now what we are talking about is, okay, you are already tested once. Is there an added value of questionnaires. That is really what we are trying to get at. And, unfortunately--and you asked the right question but, unfortunately, there are no data to tell us how often, or with what likelihood, we would pick up a recent infection. We just know it happens. But we don't know how often and we didn't see it.

DR. FORSHEE: And if I may say just one more thing about the modeling. Some of you have a very good intuition there are other ways the model could have been set up to look at process of doing a questionnaire, following them up for repeated testing. There are other ways the model could have been set up. But what we decided to do was to use the empirical data that we had showing the relative sensitivity of test-once, test-twice and universal. We decided that, out of all the choices we had about tradeoffs between the quality of data, we decided that that was the best data that we had at this moment to show the relative effectiveness of the three strategies.

DR. SIEGAL: Dr. Cryer.

DR. CRYER: Unfortunately, that doesn't help us answer Question No. 1. But we also do have data here--I mean, real data--that says that if you ask people the first time, before you tested them, that they get it wrong; right? So one would presume that there would be no different pressures involved or any different insights involved when they answered the question again. So it is an assumption, a supposition. But at least we have data for the first half.

So it seemed to me that this question for me is pretty simple. I mean, they are going to answer wrong again.

DR. McCOMAS: If I may follow up on that just to confirm that research, on-survey research, indeed shows that there is a self-report bias so, from other studies, people are going to get it wrong, not necessarily that they mean to lie but they misunderstand the question or they don't really understand the reason behind the question.

So, whereas, I don't think that questioning is necessarily a bad thing to do, I think that these questions appear not to be the right ones to ask or they are not being asked in the right way. And, if the FDA or the blood industry wants to have questions, and I perceive that they might, that they really ought to invest in a little bit of analysis to understand why people were systematically answering it wrong and what are some of the built-in procedures that might be influencing those wrong answers.

DR. Di BISCEGLIE: I would suspect that the real meat, the real discussion, will be in Question No. 2. I have a sense that the committee knows what the answer is to Question No. 1 and we might want to get on to Question No. 2.

DR. SIEGAL: I think that is maybe a good suggestion and, perhaps, we ought to take a vote on this. Are people willing at this moment to do that? Dr. Hollinger?

DR. HOLLINGER: I still have a problem with this question, also. And it also is tied into the others. I think this is what the issue is. You are presented with two or three alternatives, the optimal of which is universal testing. And then we talked about one-time testing, two-times testing, and we talked about testing once and then maybe five years later which is probably a very difficult thing to do.

And then we talked about the question and so on. But, I think there is a lot of good data, it seems like here, that selective testing can be very useful. However, under the circumstance of not knowing for sure about incident cases, both in this country as well as abroad--I think Dr. Stramer's data was that only about a quarter--about a quarter of the RIPA-positives were autochthonous, or may be autochthonous. The majority were from outside. I think it was 40 of the 157.

So most of the others were presumably from acquisition outside the United States. So the question that was not asked, to my satisfaction, was, since your last donation, has this happened. Now, granted, there are not very many incident cases in that 22 months, which is a short period of time. I am much more comfortable, personally much more comfortable, until we know this data and this data is available to us, to have a question like that asked of some sort, particularly the one about, have you been out, say, for more than a month south of the border, south of the U.S.-Mexico border, any travels down through Latin America or South America including the Caribbean.

Until we are sure that, with the data, that there are no incident cases occurring in this group that we were going to let--that would not test. After several years, we would have that data and then we could make a better decision about how to proceed in there. I just think it has been too short to come to the conclusion that just selective testing once is sufficient.

DR. BIANCO: I think, Blaine, that this is proportional to the size of the population of positive donors that we are finding in this experience of two years. How much are you going to wait by that strategy in terms of pickup.

The issue I think that we are all having--I don't have to answer the question--but that we are all having with this question is, really, does the committee agree with FDA that questions are an inadequate method to select donors for testing. This is a little bit round-about and, if it were direct and just applied to the data that we heard here, that is the answer. No.

DR. ZIMRIN: I think there are two issues. One is that questions are not a particularly good way, it seems, of getting accurate data. And the second is we don't actually really know. It is not as if, if they answered yes, we would know 100 percent sure that they would, therefore, be at risk. We don't even really know what the risk factors are. So not only is the method not good, but we are probably not even asking questions that are going to give us a reasonable answer.

So I would suggest that, until we have data that suggests that questioning really does some good, it would not be a good idea to mandate its use.

DR. SIEGAL: Dr. Colvin.

DR. COLVIN: It seemed like there was one question that was a problem and that was the question as to where somebody was born. There could be many reasons why that question could be answered incorrectly that would probably be purposeful by the person answering that question. In this case, a lot of times, they may be afraid of immigration status and what somebody might say or suggest about their immigration status depending on where they were born.

The other questions, it didn't seem like they were that far off. There weren't different answers every time they were answered. So it seemed like one particular question. Now, the question you would ask on secondary testing would not be where you were born anymore, because there was already a negative test. The secondary question would be, since the last time you donated, have you spent a month overseas or in Mexico or in South America, which is a very different thing, I think.

Obviously, it is a hypothesis that I cannot prove that it has something to do with immigration, but it is one possibility.

DR. BIANCO: I think that you are correct, that this has biased many of the questions. But Dr. Maguire just told us that people have to live in those houses for years in order to develop an infection. It is not something, except for the couple of weird incidences of oral transmission that happened in the last few years because they ingested the bugs, the mashed bugs, as they prepared juices and all that.

The possibility of getting infected just by the housing and all that is minuscule in travel.

DR. MAGUIRE: [Off mike] --an area where the transmission--you have got the control measure. I mean, there are still parts of Central American, northern South America, where control measures aren't that far. But, still, the risk of getting infected--let's say you spend a week in such a house--is tiny.

DR. SIEGAL: Anyone else? Then let's proceed to a vote on this question.

DR. FREAS: Before you vote, please double-check to make sure you picked up your correct remote. There are 17 voting people at the table. The industry rep is not voting. When everybody has voted, please show the results.

[Simultaneous voting.]

DR. FREAS: Dr. Nelson, Dr. Colvin, Dr. Glynn, Dr. Maguire, Dr. Trunkey, Dr. Zimrin all voted yes. Dr. Hollinger voted no. Dr. Bower voted yes. Dr. Siegal, Dr. Ballow, Dr. Blackwelder, Dr. McComas, Dr. Cryer all voted yes. Dr. Kulkarni, Dr. Di Bisceglie, Dr. Rentas and Dr. Finnegan all voted yes. So it was all yes with 1 no. That is 16 yesses and 1 no.

DR. SIEGAL: So let's proceed to Question 2A, which also requires a vote. I can read the question; do the combined scientific data on risk of transfusion transmission of T. cruzi support a selective-testing strategy in which, A, one negative test would qualify a donor for all future donations without further testing or questions regarding risk of a newly acquired infection? Discussion?

DR. TRUNKEY: I would ask Dr. Epstein a question in regards to the immune-incompetent individual. Would you use a one-test or two-test if you identified a potential donor for solid organ or tissue?

DR. EPSTEIN: Well, I don't know the answer. We are here to be advised. I think you did bring up earlier the question of whether you would want a more scrupulously tested donor pool for your most immune-compromised recipients. I don't think that we have really had a full deliberation. I think that the general framework is that we set a safety standard and we screen donors according to that standard.

Now, there is the model for CMV where we do not have a routine requirement to test but where we do have approved screening tests and where discretion is exercised by blood banks to provide CMV-negative blood for at-risk recipients. So I think we could think in those terms if we believe that additional testing adds value.

And, anyway, even if we recommended one-time testing routinely, it could still be a industry practice voluntarily to screen twice for an immune-compromised recipient if there weren't otherwise a consensus.

DR. TRUNKEY: I am not sure I have the knowledge to ask this question, but it seems to me there is some value in using two tests which increases the rate, if you will, of knowing whether a person has this particular antigen; is that correct?

DR. EPSTEIN: Yes; the data do indicate that at least two-times testing improves sensitivity about 95 percent up to the 98 percent range.

DR. TRUNKEY: But, unfortunately, Dr. Montgomery did not give us the denominator of the people. But, it seems to me, if there were five cases of Chagas in donors and it went to 19 people who got infected, that seems to me an increased risk factor. I may be all wet on this. I don't know.

DR. BIANCO: Are we discussing--I think the question from Dr. Trunkey is very important, but are we discussing blood donors or are we discussing organ donors and recipients. They are very different.

DR. EPSTEIN: That's true. But I think if I understand Dr. Trunkey correctly, and you can rephrase it if I am wrong, you are asking whether we should entertain a concept of a twice-tested donor selectively for an at-risk recipient for an immune-compromised recipient, especially an organ recipient but perhaps others as well.

We are talking about screening blood donors. We are not talking about screening organ donors although the same concepts might apply.

DR. BIANCO: Dr. Epstein, these may apply very well for, for instance, cadaveric specimens. They are much more difficult to test. But you don't have two specimens in a cadaveric donor to test. You have one.

DR. CRYER: I think it is more about the recipient. So, in other words, a bone-marrow transplant patient is going to be getting a lot of blood in our institutions as are all the other transplant populations. So the blood supply that is coming in for that group of patients potentially would need to be safer than it is for people who are not immune-compromised that are going to receive blood. So it is about the recipient, not the donor, not the donor of the organ that we are testing. It is the donor of the blood for blood that is going to be transfused into immune-compromised--

DR. BIANCO: Yes, but it will create, Dr. Cryer, a complexity that is very--requires serious thought because you will have to create a population of donors that you maintain an inventory on the shelf that have been tested twice to be provided to those recipients. So you have to balance the complexity, the possibilities of error and the fact that people may not be necessarily--the physician may not necessarily request that unit at that time.

DR. ZIMRIN: I just wanted to say that I thought the doctor from Texas, who spoke to us, made an excellent point. I think that one thing we sort of lose track of is that we have moved, over the years, from a discussion about HIV and hepatitis, which caused a lot of mortality and certainly a lot of morbidity, to West Nile where there was clearly an infectious disease that caused morbidity and some mortality as well to something where we don't know the clinical implications, where I think there is a lot of uncertainty.

I think this is, again, a test that we have available but trying to make decisions as to how much resources to commit to this, trying to make a risk/benefit analysis when we don't really seem to have a good handle on the risk, or we at least don't know the lower level--well, we really don't seem to have a good idea of what it is. It certainly doesn't seem to be as high as some of the other diseases that we are comparing it to.

So I just think we need to keep that in mind as part of this discussion when we are thinking about what kind of mandates we are going to recommend.

DR. BALLOW: To test once seems to pick up the individuals that were born elsewhere. And we get a 95 percent exclusion of those donors. Doing two tests increases it to 98 percent which is pretty close to universal testing so I think probably people around the table would say there is no sense doing universal, that testing twice gets us almost at the same point.

But the real black box is the autochthonous individuals who acquired this disease, or at least acquired infectivity, who were not born in South America or Mexico. That is the information we don't know. Apparently, the CDC is looking into further--trying to get some further information about those individuals and how they acquired the infectivity and what the clinical consequences are.

So, without that data, it is really difficult to decide whether we should test once or test twice since we don't know what the clinical outcome of those individuals are. So I don't know whether we would be conservative and just say, okay, let's test twice and, hopefully, in two years, we may come back and we will have that data about those individuals that acquired the disease in the United States and then may have to change our policy. I mean, that would be a conservative approach with those individuals.

DR. SIEGAL: Dr. Cryer.

DR. CRYER: I see it a little bit differently. So, if I do the math right, if you test once, you are going to have 20 patients left in ten years that are going to get the disease from the blood transfusion. If you test twice, you get rid of 12 of those, so you are down to seven. And then, if you do universal, you knock it down another six, down to just one. So, if you are going to spend the money to knock it down 12, then it ought to be worth the money to knock it down another 6.

The trouble is, if I heard the math right, the cost of that at $100 million a year over ten years for knocking 19 patients out of the pool is $1 billion.

DR. GLYNN: I think I agree that the issue is the incidence because we only have data for about 10 months on the incidence, so the incidence is zero over 10 months. But we don't know what is going to happen over the next few years. And the way those questions are phrased, it looks like you test once or twice and then, after that, you don't ever test them again. So is that correct? Or is there another alternative?

DR. DUNCAN: That is the correct formulation of the strategies that are in these questions. The concept has been raised of testing again after two years, five years, and we would entertain comments from the committee on Question 3 for those kind of suggestions.

DR. EPSTEIN: I just want to clarify one point which may or may not be clear to everyone. A two-times testing strategies is not the same thing as saying retest in five years or retest in ten years. A two-times testing strategy could be that the second test happens at the next donation in 56 days, or it could be it happens in 48 hours at the next platelet donation. It is simply twice-tested. It is not an interval test inherently.

Now, on average it is, because, on average a repeat donor gives twice in three years--I'm sorry; three times in two years, 1.5 times per year. But, bear in mind, that is a short interval. In the actual study, the average was about 8 months. So, just to be clear, when we talk about a two-times test, we could be talking about two different things. We could be talking about a retest at some fixed interval or we are simply talking about test-twice.

Now, the test-twice argument had nothing to do with picking up incident infections. It only had to do with picking up some proportion of the true positives that are fluctuating near the assay cutoff because the antibody titers are low. So think of that more as replicate testing then interval testing. They are really two completely different ideas.

DR. GLYNN: But, Jay, just to clarify, the one we are asking to vote on is that one, the second one, the repeat testing.

DR. EPSTEIN: Well, yes. Replicate testing. In other words, on more than one occasion but without any specified interval. And that concept is targeted toward picking up again some proportion of the low-level reactives that are true positive but might be missed on any one-time assay because they are so near the cutoff, just outside variation.

DR. Di BISCEGLIE: So that is useful, Jay. I mean, there are these two thoughts, picking up new cases, incident cases but I think, more important, trying to confirm the negativity by basically repeat testing. It seems to me like the science is a little bit in evolution I mean, this is firstly a new--I am not sure of the right work, mandate recommendation, from the agency.

The test is fairly new and we really don't have adequate confidence intervals around that rate of incident cases of zero. So I wonder if a suitable strategy may not be to say we have done a good thing by doing one test and now we need to study it further to look at incident cases, whether questions can identify them, what interval it might take for new cases to appear. Nine months may not be enough. It may be five years, and isn't it likely that a newer and better test will emerge. It is the nature of things, compared with competition, that newer and better tests tend to emerge.

DR. BIANCO: And adding to the points that Dr. Di Bisceglie made and Dr. Epstein, maybe we should consider the consequences of another change in cutoff. I don't know. I know Dr. Stramer would hate that. But, in a certain way, if the impact is small, or tolerable--I wouldn't say small--it may be easier to deal with than repeat tests.

DR. STRAMER: Let me just say it is a slippery slope. It is never going to end because of the dynamics of the test. We will always find, then, other samples that are below the cutoff, so it will never end and we will negate all the work that Ortho did in design of experiments to make this test highly specific. We will chew away at all of the confidence that we have that the test is specific.

DR. BIANCO: Aren't you going to have the same problem multiplied by 2 if you test twice?

DR. STRAMER: No.

DR. CRYER: That question for me is not so important as it is how many--if you went universal. So the difference how many tests you would decrease by doing it twice from doing it universally. Do we have an estimate of what that might be? It is easy from 1 to 2. It is half. But what is not so clear is what the difference between doing it twice and doing it universally is, because that depends on how many times people come back.

DR. BIANCO: Dr. Busch has some estimate.

DR. BUSCH: I showed those numbers for the first year and a half of testing. If we went to, I think the last page of the handout--right; so in the period where we did universal testing on 1.4 million, if we went to one-time testing, it would drop to 700,000. If we required two-times testing of those donors, it would be 1 million, 4 thousand. But that will evolve over time as the repeat donors get tested once or twice.

So, eventually, theoretically, with a one-time donor strategy, we should settle out continuing to need to test about 22, 23 percent of donations because these are the first-time donors coming in. If you have to test all donors twice, you know, it would probably settle out at 25 to 30 percent of donations will be need to be tested on an ongoing basis.

DR. NAKHASI: Hira Nakhasi. I just wanted to emphasize the point there that Celso was sort of suggesting and I guess Sue answered to some extent, it is not only reducing the specificity and increasing the sensitivity, you will increase the false-positives and we do not have a supplemental test. Therefore, you are going to increase more people false-positives which are varied well, and that is another consequence of that.

DR. ZIMRIN: It seems to me the most clear improvement we can make over the strategy of doing nothing is to test once. After that, the improvements are relatively small and the gain is relatively ill-defined. So why not recommend something that seems likely to be a positive and continue to investigate until we actually know more about what are recommending.

DR. DUNCAN: One of the points I wanted to make about the question of incident infections, whether it is from travel or whether it is from autochthonous transmission, is, if that is still a concern and we still need to do more study, then what testing strategy we propose will affect that greatly because, at this point, the only testing strategy that will capture true seroconversions, true new infections, would be universal testing or maybe some sort of time-dependent multiple testing short of universal testing.

Whether we test once or twice, anything else would be a study that would be outside of the standard blood-donor screening.

DR. ZIMRIN: Are you saying we can't study this unless we mandate it. That is the only way to get the information.

DR. DUNCAN: I would not say we can't study it, obviously. There could be ways, but the kind of studies that the Red Cross has done, that Blood Systems, Incorporated, have done, where they are doing the study in the context of donor screening are just more productive because of the sample size they are able to marshall.

DR. BUSCH: I would suggest a study could be done. And, because we are so in favor of moving to selective testing, that, if a kind of postmarket--so, let's say you guys approved one-time testing but with the understanding that a study would be conducted perhaps in two or three years where we would retest, perhaps, in some of the higher prevalence regions, the regions in which there are higher rates of persons from those countries.

We could test 200,000 additional donations from donors who had been tested once before who have come back now, two, three, four years later, so some kind of postmarketing study could be conducted in a few years to just verify that we haven't seen incident infections. We could also continue to enroll and follow and get serial data from the positives which is, again, a very good approach to identifying recent seroconverters.

So I think some sort of post-selective testing studies could be designed and we would make the commitment to conduct those downstream to verify absence of seroconversion.

DR. GLYNN: I agree that that makes a lot of sense to me. It is more reassuring to know that later on you would have data that hopefully would confirm your original decision because, otherwise, again, the way it is worded right now it is like you do it once and you are done forever. So I think we would have a Choice 3. I don't like A and B, myself.

DR. Di BISCEGLIE: So something that Dr. Busch said that we haven't explored very much is this question of regional variation. We did have a big discussion about sensitivity and specificity and, of course, those are very much affected by the actual prevalence of whatever it is you are testing in the particular population. So, part of further study might be the need for regional variation in the retesting strategy, perhaps, because, again, sensitivity and specificity will be different in California and Texas than it might be in Minnesota.

DR. HOLLINGER: And I think Dr. Busch's suggestion is a good one. I think you do need some additional data and it sounds like this is a disease you could probably get that in. So, doing this several years down the line, or a few years down the line, to look at--but I would also just include it over the whole group because we still want to know about incident infections in a population. That would help us a little bit understand that after several years. So, to me, that would be a better approach.

DR. SIEGAL: Are we ready to vote on Question 2A?

DR. GLYNN: Can we add that question, then, as a choice, what we just proposed?

DR. ZIMRIN: I think there is some support for that.

DR. DUNCAN: You mean the suggestion about later studies? I would suggest that we put that into Question 3.

DR. SIEGAL: There is room for it there and we can discuss that again. But let's vote on 2A.

DR. ZIMRIN: I am not clear on Question 2A. So this is regarding a mandate, not regarding a recommendation for further studies or further research. Okay. Thank you.

DR. SIEGAL: This is one negative test that would qualify a donor for all future donations without further testing or questions regarding risk of the newly acquired infection.

DR. DUNCAN: But the place it would fit in--

DR. ZIMRIN: There would be no postmarketing?

DR. HOLLINGER: According to this.

DR. NAKHASI: But then, in Question 3, you are going to [off mike].

DR. ZIMRIN: Wait, wait. Let this get real clear for me. So, if I like the postmarketing idea, I should vote yes or no for this?

DR. HOLLINGER: No.

DR. DUNCAN: If you think one test is a good idea, vote yes. You can then say postmarketing in Question 3.

DR. ZIMRIN: Okay. Thank you.

DR. GLYNN: But that is different. That is different from--the question you are asking is can you do the testing once and then, afterward, you do nothing ever again on the repeat donors. That is how I understand that question.

DR. SIEGAL: That is the question.

DR. GLYNN: That means no studies, no postmarketing studies.

DR. ZIMRIN: No; we are supposed to vote yes for this and then vote--include postmarketing in the next question.

DR. HOLLINGER: But if you answer it that way, and if you answer it the way you just stated that, that is not what it says. It says one negative test would qualify a donor for all future donations. You are not saying you could do something different. You would have to vote no on this question and then add something in the comment section about it.

DR. DUNCAN: Could I suggest we look at it this way, that this is a policy that would be applied in general to blood donors.

DR. ZIMRIN: Okay.

DR. DUNCAN: The question of another study, a subsequent study, a postmarket study, that is going to be done as a subset of people anyway. It is not something that will be applied to all donors. And the outcome of that study might change the policy that we would then apply to all donors.

DR. ZIMRIN: Okay. So this is a policy question. Okay. Got it.

DR. CRYER: Can I clarify one thing because it seems to me that, if the data model is correct, that you would have to test the entire pool of donors in a year to have a chance of just finding 1--0.7 patients--0.7 patients--that the first test was wrong in. So you are not going to do any post tests that are going to mean anything. You would have to play this out for ten years to find out whether it was working or not. And the answer to that really depends on how you would answer No. 1.

DR. DUNCAN: I would suggest that is not exactly the right interpretation. The model that says if we test once, there will be 1.9 new transmissions in the next year is based on transfusion transmissions. This postmarket study that Mike Busch has just suggested would be to try to capture incident infections that could come from vectorial transmission, that could come from travel, or other sources.

DR. BUSCH: But I think the concern is that 22 months is only less than a year in a modest number of donors. A study could be designed that would implement a question to donors who have been repeat donors and were screened negative three or four years ago, and now we are going to target donors who may have traveled outside the country, so we could add a special question, or donors who live in the high autochthonous zones, and test half a million.

I mean, it sounds like a big number, but, for blood screening, that is trivial. And the amount of money that we will save over the course of three or four years of single-donation testing, it is well worth a large study to endorse and determine whether this strategy is safe.

DR. HOLLINGER: Jay, this is a stand-alone question.

DR. EPSTEIN: Yes.

DR. HOLLINGER: And, if voted on as a stand-alone question, it opens itself up to a lot of interpretations. And you just heard the discussion around the group here, because the way it is written is very clear. It doesn't say anything about doing something later on, considering something later on. It says, specifically, one-time testing. One negative test would qualify a donor for all future donations without further testing or questions regarding risk. It doesn't say about any further testing down the line.

So either you have to add something at the end of this that says, "although something else will done," or, "we will consider other possibilities," or you leave the question like this. If you do, and somebody wants to do, or wants to consider, some repeat testing in a certain period of time later on, they would have to answer "no" to this question and then put it in the comments afterwards.

DR. GLYNN: I agree. That is my interpretation as well. I mean, I agree completely with what you said, Blaine. That is also what I understand from the question.

DR. EPSTEIN: I think that it is very clear that the sense of the committee, at least those who are outspoken on this point, would like to see further studies to clarify incident T. cruzi infections. I think the FDA gets that. That doesn't mean we can compel those studies. So what we are basically saying is, do you think that the data warrant a one-time testing strategy. And I think that Dr. Duncan correctly framed it. That would be the routine recommendation at present. We are not foreclosing other options in the future, but that would become the standard of practice.

So we are really saying, should we at least test each donor once. Now, you can go on and say no, but we really ought to test them twice, or you could go on and say, test them once but test them again maybe in five years. All those options for the future would be fine to recommend. But the threshold question is should we at least test every donor once.

If you think there is no value in that because of the transmission rates being so low, tell us now. If you think it has value, then you really ought to say yes even though you may have much more to say.

DR. NELSON: So, Jay, then you say the obverse of this question is not testing every donor. Is that what you are saying?

DR. EPSTEIN: If you answer no to 1 and 2, you would be saying don't test--I'm sorry; to A and B, 2A and 2B, you would be saying don't test.

DR. FINNEGAN: But 2B gives you an improved statistic.

DR. EPSTEIN: Yes; 2B is, do the combined scientific data on risk of transfusion transmission of T. cruzi support a selective testing strategy in which donors should be tested twice. But, bear in mind that that is not interval testing. It could be 48 hours later in the platelet donor. It is simply tested twice. Think of it as a repeat test.

DR. CRYER: The way I read A is it is saying is once enough, yes or no.

DR. EPSTEIN: That's correct.

DR. CRYER: Not whether you ought to do it at all, but is once enough.

DR. EPSTEIN: That's correct. But, if you answer no to A and B, you have recommended against testing unless there are other comments.

[Comments off mike.]

DR. CRYER: It seems to me that that is correct.

DR. EPSTEIN: Well, that could be the comment.

DR. BLACKWELDER: It seems to me that A is saying. is once sufficient. And B is saying, well, if you don't believe that, is twice sufficient. And if you say no to that, then--

DR. EPSTEIN: But Mike's point is correct. That is in lieu of universal testing because what we are talking about is alternative testing strategies. The baseline presumption is FDA's current guidance which is universal testing--draft guidance. What we are debating is whether there ought to be an alternative for selective testing.

DR. BLACKWELDER: But I interpret it as saying, if you say no to A, no to B, then you still have universal testing.

DR. EPSTEIN: That's correct. Unless you comment, "don't test." I was just leaving Point 3 open. You can comment back to us anything you want to comment, but the way we have framed it it is universal testing unless you answer yes either to 2A or 2B as the committee recommendation. Again, that doesn't foreclose other advice. I think we have already had the sense of the committee that, because there is concern about incident infections, there is a need to continue to study that possibility.

But, short of recommending universal testing, it won't happen automatically. I mean, somebody has to want to do the study, I heard Mike was interested, but, you know, we can't mandate that.

DR. BOWER: It could also be what was mentioned before about testing every one to two or five years. I mean, that could be part of the comment and that would be something that could look into incident infections outside of postmarket; correct?

DR. EPSTEIN: Yes.

DR. BOWER: Thank you.

DR. HOLLINGER: But, Jay, I think this is where the FDA has gotten in trouble before about not mandating postmarketing tests. I have said in these committees before about doing some postmarketing thing and then the company never follows through on it or often does not. And nobody sits down and says, you need to do this.

So I think that is a dangerous slope to be on. I think it has to have some follow up. I am a little concerned about that.

DR. EPSTEIN: Well, the problem here is whose responsibility. See, we have already licensed the test. And we didn't impost a postmarketing commitment at the time of licensure. So, generally speaking, unless there is a failure of a product, we don't go and mandate it after the fact. So the question is who carries the burden.

What we are talking about is the burden to the blood operators. And I am not sure what the regulatory framework is by which we would mandate that. We could recommend the tests in three years or Q-three years or Q-five years as guidance. We could certainly do that. But I think it is outside the framework of what was the condition of licensure. That is all I am saying.

DR. ZIMRIN: One option would be to recommend one test, or to make policies that a single test--and we are operating under the assumption that a further study will take place that will clarify the issue. If, four years from now, the study hasn't taken place and there is no evidence that anyone is making any effort to make it take place, a future BCAP committee might, then, say that the committee did not get the results they expected, we no longer feel comfortable with continuing with one test because there is no follow-up data, so maybe they ought to now move to making it more formal, that an every-five-year or two tests, or whatever they want to do at that point so it is not that we don't have options or someone doesn't have options down the road.

DR. SIEGAL: I just don't see how we can vote yes on this because forever is a long, long time. And, if we agree to this per se, then we are stuck with it, it seems to me.

DR. Di BISCEGLIE: I would suggest we are stuck with it until data emerge that would motivate for a change in policy, perhaps,

DR. SIEGAL: Yes, but where are those data going to come from?

DR. DUNCAN: I would like to interject, also, that, since I worded these questions, that the sense of the all-future donations without further testing was to make clear the nature of the testing, not the nature of the policy. The policy could change. We are not saying yes to this question. It doesn't mean you approve that policy forever. That just means you approve of that testing strategy now.

DR. EPSTEIN: Could I suggest that, if we reword the question, subject to the understanding that additional studies will be done to investigate the incidence of new infections in donors, would one test negative qualify a donor, da, da, da, da, because I think what the committee is reaching for is to link these two ideas and not very happy postponing it to Question 3.

So, if we simply link it, then I think we may have a question that the committee is prepared to vote.

DR. SIEGAL: So, can you guys rephrase the question?

DR. EPSTEIN: Yes. Subject to continuation of studies on incident infections--well, that should be the first--well, either way; yes.

DR. DUNCAN: Subject to what?

DR. EPSTEIN: Subject to continuation of studies to define the incidence of new infections in donors--new infections in previously screened negative donors.

DR. SIEGAL: Thank you. Does that satisfy the group?

DR. BLACKWELDER: Well, I'm sorry; not quite.

DR. SIEGAL: Okay. Well, that is why I asked.

DR. BLACKWELDER: My understanding, and I am repeating, I believe, that the difference between one negative and two negatives, in the modeling anyway, was based strictly on sensitivity of the test. I had nothing to do with incidence of new infections.

DR. FINNEGAN: I would suggest that that same line needs to be added to B because we don't know enough about the disease. So it really needs to be added to both of them.

DR. DUNCAN: Right; and that is absolutely what I was intending to do because, as Jay said earlier, the one-test/two-test strategy is about more sensitivity, not about capturing new infections at all.

DR. SIEGAL: All right. So, shall we vote? Yes. Let's proceed. There is a question in the back, a couple of questions in the back.

DR. BENJAMIN: Richard Benjamin, Chief Medical Officer for the American Red Cross. I work for the Red Cross. I don't have any conflicts. I wanted to speak for the American Red Cross. We strongly support the suggestion that has been put forward by Mike Busch that we have committed ourselves to going to single testing but we would need to sit down with the FDA before we actually implemented such as test anyway and, at that time, our organizations, and I believe Blood Systems will make the same commitment, would commit to doing a follow-up study at an agreed time with the FDA to look for incidence infections.

[Seconded.]

DR. SIEGAL: Thank you. Anybody else?

[Comments off mike.]

DR. FREAS: All votes are zeroed out. We are ready to vote. Go ahead and vote. We are voting on 2A at this time, only. I believe there are still 17 voting people at the table.

DR. SIEGAL: Where is 2A on this thing? There is 1A, 2B, 2C.

DR. FREAS: You are voting on Question 2A at this time. If you push 1, it means yes. If you push 2, it means no. If you push 3, it means abstain.

[Simultaneous voting.]

DR. FREAS: Dr. Nelson voted no. Dr. Colvin, Dr. Glynn, Dr. Maguire, Dr. Trunkey, Dr. Zimrin, Dr. Hollinger all voted yes. Dr. Bower voted no. Dr. Siegal, Dr. Ballow voted yes. Dr. Blackwelder, Dr. McComas voted no. Dr. Cryer, Dr. Kulkarni, Dr. Di Bisceglie, Dr. Rentas voted yes. And Dr. Finnegan voted no.

I count 5 nos, so, there are hopefully, 12 yeses.

DR. SIEGAL: All right. Then we need not vote on Question B. I think that is true, isn't it? Do you want us to vote on B also? No.

DR. SIEGAL: So let's proceed to any further discussion on 3 for which there is no vote requested. Is there any further discussion?

DR. BOWER: I will go ahead and say that--what I think the committee has talked about is that, to look into the question of incident infection, that there should be some postmarketing test or that the FDA should require or recommend that they do testing every two, three, five years, some number to be determined, to look into the question of incident infection in the U.S.

DR. NELSON: I would also think that, as Dr. Di Bisceglie said, that we might focus some prospective studies on donors from California, Texas, Florida. And the other issue, very minor but theoretical issue, is that it isn't sensitivity and specificity that changes with prevalence but it is predictive value, positive and negative predictive is a fundamental epidemiologic concept.

DR. SIEGAL: Any other commentary? Well, with that in mind, we will call this session to a close and reconvene tomorrow morning.

[At 6:45 p.m., the meeting was recessed, to be resumed the following day at 8:00 a.m.]